Abstract
To distinguish colon from period is essential to modern metrical studies. Cola are in synaphea while the end of a period is marked by pause which is sometimes, though not always, explicitly indicated by hiatus/brevis in longo. But ancient scholarship took all divided lines as equal entities and did not care about the difference between colon-ends and period-ends (in the modern sense) : 1. Hiatus/brevis is located within a line by the colometry of papyri(Pindar, Paean 2. 25, 77; 6. 16, (?)95, (?)136; Bacchylides 10. 15, 33, 43) and scholia vetera of Pindar. 2. Hephaestion regarded each final ουλλαβη as αδιαφορο&b.sigmav;, and sometimes interpreted as αδιαφορο&b.sigmav; a real long syllable which was situated at a short element(e. g. -〓〓--〓- as achoriambic catalectic(=-〓〓--〓〓〓)or-〓-〓-- as τροχαικον ημιολιον(=-〓-〓-〓〓〓))- The Pindaric scholiast even supposed such an aαδιαφορο&b.sigmav; could be in synaphea with the next line (e. g. Pyth. 8. 24 θιγοσα νασο&b.sigmav; τε-, which was described as iambic dimeter brachycatalectic). 3. Horace carefully imitated his Greek models as represented in Hellenistic-Roman editions, but failed to notice that hiatus/brevis and synartesis were incompatible in the same position. He admitted both at the end of the third line of the sapphic and the alcaic stanzas (hiatus/brevis being rejected by Sappho and Alcaeus, as the third and fourth lines belonged to the same period). 4. Horace also imitated apparent interlinear hiatus/brevis of epodi of Archilochus faithfully. Actually Archilochean models must have comprised three periods though they were divided into two lines (cf. Cologne Archilochus) , perhaps to save wasted spaces in a papyrus roll. Hephaestion seems to have developed his theory of asynarteta at Ench, 15. 1 by observing the line division of a Hellenistic-Roman edition. He did not mean a verse (a period, in the modern sense)by στιχο&b.sigmav; but a line which he found written down continuously (cf. his definitions of στιχο&b.sigmav; and κωλον at Poem. 1 and usages of these words in Heliodorus). The analysis into metra of Hephaestion and Pindaric scholia sometimes approaches nonsense. They cut a ready-made line into metra arithmetically from its beginning and called it hypercatalectic etc. paying no attention to its nature. Sometimes a right colometry was acquired. For example, the glyconic was picked up in the papyri of Sappho 96, Pindar Parth. 2 and Bacch. 2 in spite of synartesis. But in other similar odes a wrong colometry is offered : see Alcaeus 130, 357, Bacch. 6, 18. The basic principles of Hellenistic-Roman colometry are (1) to avoid split words as much as possible(2)to make each line of roughly equal length (i.e. with variation in length no greater than dimeter to tetrameter) (3) to arrange consecutive lines of similar shapes when possible. A right colometry and a reasonable classification of cola are induced from collecting all possible parallels. To decide which form of the colon is regular, the statistical method is useful, and antistrophic responsion is most important. 'Genealogical' speculation must be avoided at this stage. A colometry is argued rationally though the choice between two possible colometries is sometimes a matter of taste and though not every structure of a period can be clearly analysed by dividing into cola when the length of a colon is presupposed.
